THE STRUCTURE OF METEORITES I 89 



seemed analogous in his opinion to the needles formed on the 

 surface of freezing liquids, while the grains seen in meteorites 

 seem more like the forms of frost or snow formed by the 

 immediate passage of water vapor to the solid state. This con- 

 clusion led to a number of experiments by Meunier with vapors 

 of silica, magnesia, etc., for the purpose of forming in this man- 

 ner if possible mineral aggregates resembling the meteorites 

 The results so far obtained have been chiefly negative, though 

 the experiments are still being conducted, A prominent feature 

 lacking from the results of the experiments with vapors is the 

 glass so abundant in meteorites. This is obtained in quantity 

 however in the products cooled from fusion. 



Many efforts have been made to reproduce Widmanstatten 

 figures but they cannot be said to have met with much success. 

 As it is held that the Widmanstatten figures indicate slow cool- 

 ing from fusion, Sorby fused together iron, nickel, and other 

 constituents of an iron meteorite, and allowed the mass to cool 

 very slowly. On examining an etched surface with a lens, min- 

 ute lines were seen which recalled the Widmanstatten figures 

 but the appearance as a whole, he states, was very different. 

 The mass was then kept for a long time at a temperature just 

 below that of fusion but the. resulting product was less like 

 meteoric iron than that previously obtained. Daubree fused a 

 portion of the Caille meteorite in a crucible of clay, then allowed 

 the mass to cool slowly. The resultant mass showed a crystal- 

 line structure but all trace of the former Widmanstatten figures 

 was lost. On fusing a mixture of iron, nickel, troilite, and 

 silica, he obtained a product showing dendritic figures. On 

 fusing a mass of iron, nickel, and iron phosphate together, a 

 mass having a reticulated structure with angles showing dodec- 

 ahedral crystallization was obtained, The structure of this 

 mass approached more nearly the Widmanstatten figures than 

 any obtained in any other way, and it is sometimes stated 

 that true Widmanstatten figures were produced. As Cohen 

 remarks, however, the statement needs to be supported by fig- 

 ures and specimens. The same may be said of Meunier's 



